search

informatics-isi-edu / microscopy

A microscopy image management solution based on the ERMrest service.
Apache License 2.0
1 stars 1 forks source link

Not getting whole czi image in pyramid #34

Open ljpearlman opened 6 years ago

ljpearlman commented 6 years ago

Jill McMahon raised the issue that for https://www.rebuildingakidney.org/chaise/record/#2/Gene_Expression:Image/RID=14-47WE, is only showing a portion of what she's expecting to see. I tried downloading the czi file and opening it in Fiji, and it looks like Jill is correct.

macro_image

hongsudt commented 6 years ago

@ljpearlman is it just one that has this problem, or there are more than 1? I wonder whether they can rescan.

ljpearlman commented 6 years ago

Jill noticed this because there was an oGUDMAP thumbnail that didn't match the viewer image, so there may or may not be others she hasn't noticed. It looks like this image was originally uploaded to CIRM in December 2014, so I don't know how practical it would be to rescan.

hongsudt commented 6 years ago

The image just looks kind of terrible to me somehow..

ljpearlman commented 6 years ago

Oh - that's just one layer. I just included it to show that there are supposed to be three pictures on it, not just the one in the center. If you follow the viewer link, the quality (of the part that's visible) is comparable to a lot of the other stuff in GUDMAP.

svoinea commented 6 years ago

There are no differences in the way that czi file was processed on CIRM and GUDMAP. Therefore, it is not practical to rescan it. On CIRM, the URL is https://cirm.isrd.isi.edu/chaise/viewer/#1/Microscopy:Scan/id=947

hongsudt commented 6 years ago

I can't see the cirm image.. Does CIRM have the same problem as what we saw in GUDMAP?

From: svoinea notifications@github.com Sent: Wednesday, October 31, 2018 5:57 PM To: informatics-isi-edu/microscopy Cc: Hongsuda Tangmunarunkit; Comment Subject: Re: [informatics-isi-edu/microscopy] Not getting whole czi image in pyramid (#34)

There are no differences in the way that czi file was processed on CIRM and GUDMAP. Therefore, it is not practical to rescan it. On CIRM, the URL is https://cirm.isrd.isi.edu/chaise/viewer/#1/Microscopy:Scan/id=947

— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/informatics-isi-edu/microscopy/issues/34#issuecomment-434895856, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AAZUDEY0QY6HlpFnerfl11Ip-0gvCI00ks5uqkbkgaJpZM4XaJzj.

svoinea commented 6 years ago

That is what I mentioned in my previous comment.

svoinea commented 6 years ago

By adjusting some values in the generated ImageProperties.xml file by our czi processing tool, there is now an image as expected. I have set the minLevel to 1 (was 0) and increased by 20% the width value.

If you click the Zoom in button, the first time it will show you a black image (due probably to the fact that the minLevel was already increased by 1), but if you click it again, it will show you the magnification of the image increased twice.

An alternative should be to leave the ImageProperties.xml file unmodified. In that case you don't get all the three kidneys on a single image, but after you Zoom in, you can shift the image left/right and they will appear.

No modification was done in the czi processing tool; just in the resulted ImageProperties.xml file (in case you want to have all the three kidneys simultaneously visible).

For this particular czi file, there are only 7 zoom levels (most of the czi files have 8), but so far, I can not see anything wrong in our czi processing tool (the tiles are generated; the problem is the visibility).

svoinea commented 5 years ago

In the ImageProperties.xml file I have set the maxLevel to 5 (was 6) and renamed the tiles directories from n to n-1.

The image (https://www.rebuildingakidney.org/chaise/record/#2/Gene_Expression:Image/RID=14-47WE) looks good now and the zoom is working as expected.